Edits to fix up this game

[RAVENOUS]

Quote: Originally Posted by Mick-S

not sure doing it by player ratings would be any better, the team with the best players will always win, at least there way you get some upsets, or maybe it was just to complicated to do another way, thing is EA guy`s are not stupid and I am sure if there was a better way they would have used it, but who knows, thing is it`s the only NFL game we got.

Hmm? You could definitely make it so the team with better players was more likely, but not guaranteed, to win. Also, I think most of the Madden players out there would say the Madden developers did most things in the game the lazy way and not the "better or best" way.

What we need to figure out is 1) If Madden generates a list of possible outcomes and then picks from one and 2) When this is done, so we can hopefully manipulate ratings and/or sliders to make the outcomes more realistic.

Well, in addition to so many other things.

[trey31]

Quote: Originally Posted by Mick-S

well you need the EA DB Editor just in case get it here : http://www.mediafire.com/download/xp...B%20Editor.rar a dynasty file already de-crypted https://www.mediafire.com/?h4iutarrj8orvof Unknowns tab/ STST is an eye opener, gives yds per carry / pass attempts etc, well the whole game player stats really.

Thanks! I'll look into this within the next day or 2. Stupid work is always getting in the way of fun...

Quote: Originally Posted by RAVENOUS

I have a franchise file with every player and coach rating set to 0, and was going to raise them to 100 one at a time and sim a season to see what ratings matter and what don't. Was just going through the traits to turn them all off/equalize them as well, which was taking awhile.

I am super curious as to how these results turn out. It will be extremely telling as to what we're going to be able to accomplish with PS3 roster/franchise editing moving forward.

Quote: Originally Posted by Mick-S

although I am not happy about it, I don`t really see how EA could do it any different.

I do. A its called an algorithm. Hell they could even adapt an open source one for free. Calculations are done in the millions per second even on single threaded CPUs from 10 years ago. They're like what google uses to tailor results from search, or what amazon uses to recommend other crap for you to buy based on your purchasing habbits, or youtube's video recommendations, which are similar to what 7zip uses to compress and decompress data without a single bit of data loss... All those would be more complicated that working out an algorithm to figure out 80-120 possible play outcomes in a football game. An entire week could probably be accurately simulated in 1-5 seconds, based on player attributes, tendencies, previous game hot or cold streaks, career stats, all weighted against the team's coaching philosophy versus the opposing coach's... There is no reason that it should not be doing this honestly. All the parameters it would need are already present.

Quote: Originally Posted by RAVENOUS

IDK myself how to actually program games, but common sense tells me there has to be an algorithm that could be run each game that would let the ratings of the players playing affect the outcome. Then again, if you look at draft classes, they just preloaded 30 seasons of those as well with no randomization other than the names, and even those weren't randomized until '12. I already had a feeling stats were awarded moreso based on who was filling what position on the depth chart, rather than who the actual players were, because I would see similarly rated WRs with the WR1 getting 85 catches per season and the WR2 only getting 55, and then WRs that were 20 points apart in overall rating with the WR1 getting 85 catches per season, and the WR2 only getting 55. At the very least, ffs, if they were gonna be so lazy about it, they could've predetermined some realistic season totals. Never see a QB top 63% completion, and the top 10 are usually clustered between 58-61%.

This. Exactly this.

Quote: Originally Posted by Mick-S

...but who knows, thing is it`s the only NFL game we got.

And that's the underlying cause of this whole thing...

Quote: Originally Posted by Mick-S

well they may well be, because I assume after week1 is finished it generates the week2 results, how it determines this I have no idea, but your theory makes sense. might be worth starting a franchise sim a few games, then start another franchise with the same roster and settings and sim the same games and see what scores you get.

Yes. I'm curious to see what happens when everyone is rated a 0 as well and how it affects this.

Quote: Originally Posted by RAVENOUS

Hmm? You could definitely make it so the team with better players was more likely, but not guaranteed, to win. What we need to figure out is 1) If Madden generates a list of possible outcomes and then picks from one and 2) When this is done, so we can hopefully manipulate ratings and/or sliders to make the outcomes more realistic.

Agreed. Even if its half baked, it could still be fixable.

[RAZRr1275]

I can do some testing starting tomorrow if you like. Just let me know what to test and which file to use.

[kcsam]

Quote: Originally Posted by RAZRr1275

I can do some testing starting tomorrow if you like. Just let me know what to test and which file to use.

Same goes for me. I'll have more time Saturday, but I may be able to fit in some testing time late Friday night.

[trey31]

I posted test roster #3 in the Slider thread. Also I outlined all of the changes and posted img captures of the excel sheets.

[goravens2052]

So far I have only done the first two things, pass blocking ratings up to 99 for all interior lineman and the good DTs having the moves in order to be able to do their thing.

First game I tried(PS3), playing as the Ravens against the Patriots and first sack of the game came from DE Chris Canty. Linebackers were able to get to the quarterback much better, and more realistically like you see in the NFL. One play, Suggs came untouched to Brady for a sack. Have to lower the jump on LBs as the Super LB came into play for an interception that should have been a TD.

Loving the improvements or edits as the title says. Great job!

[RAVENOUS]

Been away for a couple days, gonna attempt to get back to some nitty-gritty. I still need to set all the traits to be the same for every player, but it takes so long that I may just load up the season with players and coaches at 0 ratings at least and see what happens. I'm hoping to get to a point where (nearly) every single team finishes 8-8, and every player has the same statistical output for their position, but I do think traits are going to need to be leveled out for that to happen.

EDIT: one thing I'm noticing, whether or not it makes any difference is yet to be seen, is that every single non-QB has "Forces Passes: Aggressive" and "Sense Pressure: Trigger Happy." Every single offensive player "Plays Ball in Air: Balanced," only a few secondary players have "Aggressive" for this trait, and a few non-secondary defenders have "Conservative."

[RAVENOUS]

Quote: Originally Posted by trey31

If I start a franchise in week 10, I accurately sim the first 10 games, not to accurate scores, just Win or Loss...)

Responding to this a few pages ago, the "list of possible outcomes" might not be an entirely awful thing for purposes of what you're saying here. If we could go in and find this "list" in an editor the same way it can be done for NCAA, we could use the editor to make every single possible outcome be the one that actually happened, and have not only W/L but every single statistic correct for starting a Franchise in week 10, or progressing a '12 franchise file to current day.

I know the file I had last year had the '13 schedule in addition to players, so I'm hopeful...

EDIT: Something else just occurred to me on the predetermination tip: there have been games I really did not want to play in the rain, and no matter how many times I exited/restarted the game, even reload the franchise or restart my Xbox I could not get the game to be played in decent weather....